State Of Match Making - Feedback/comments
#901
Posted 08 January 2016 - 11:08 PM
#902
Posted 09 January 2016 - 06:23 AM
SgtMcFarty, on 08 January 2016 - 10:07 PM, said:
If the developer isn't tracking this data, I would be very surprised. It is not difficult to track, games do it all the time, how many people you've killed, how many shots you've fired, how long you've played, etc. For some games, a lot of this info is even available on the public profiles of players.
We're talking about different data. I should have been more clear. Not data, but data. LOL
If you read my example you will see I pointed out the complications. It is not with gathering the raw data, it's in turning that data into the type of data your suggestion requires.
Yes they track shots fired, kills, etc obviously, we see that on our profiles. They also track group size as they have mentioned it before. They can easily narrow that focus and find out how much time each pilot spends in each que....and certainly they could further query the database to spit out how much money each pilot has spent.
Again, it's the leveraging that data into the conclusions you are asking for that would be complex and conveluted to the point of uselessness like any statistic; it will say whatever they want it to. It will require interpretations and guesses and ultimately it will not do any more good than them just paying more attention to the community and raw player population numbers.
#903
Posted 09 January 2016 - 03:34 PM
PhoenixFire55, on 03 September 2015 - 01:52 PM, said:
BS.
Quality of the average match in solo queue is still p!ss poor. You still see one team with 2-3 guys doing all the work carrying 9 scrubs who barely know how to walk, while another team has 10-12 "average" players and proceeds to steamroll the former. Is that what you understand as a "quality balanced match"?
^^
bring back the old elo per mech and the old gamemode select systems for far superior enjoyable fun matches
as the solo queue is right now,it's simply a very poor match quality and not much fun
i didn't mind the wait times,for far more enjoyable matches
also CW matching pug's vs grp's didn't work before wont work ever
you already tested it,it failed
why do it again?
Edited by Fastwind, 09 January 2016 - 03:37 PM.
#904
Posted 09 January 2016 - 04:20 PM
Russ Bullock, on 03 September 2015 - 02:45 PM, said:
No point is reading and responding to your posts or suggestions because according to you we can't even tie our own shoes. I will focus on those that are more reasonable. Here in this very thread there are others that agree that the solo queue has made the best matches yet in MWO - I guess we know what you think of them.
ohhhhhooohhh this tells a lot
you did respond,i don't know if you read it
and it's nice to hear that you wanna focus on those that say what you want to hear
instead of whats rly going on
because what phoenixfire55 is discribing is pretty much what i'm experiencing too
and we are not the only ones in this thread
and the best matches yet in solo queue were with the elo system on every chasis
and the gamemode select
yes you can say the queue times are reduced now
but the quality and fun in the matches got far worse
maybe everyone who tells you otherwise just says what they are saying because they don't have to wait 4 min. to get a match
wasn't that why you changed the matchmaking because of those that couldn't wait a few minutes to get a match
you had a rly good "mech chasis and personal skill based"matchmaking system and went back to a world of tanks style non existant matchmaking system
this makes me rly sad
i guess if people are passionate about something the tone can get a bit rough
in the end who ever ownes it can do with it what he wants
if it's right or wrong,succesful or not is another story
#905
Posted 11 January 2016 - 11:39 PM
That being said, organized groups of 8+ in the public queue can unbalance the game horribly unless there is another team of similar size opposing them. This makes the game not fun.
The fact that some groups of people like to take advantage of this to the fullest extent possible is why there is so much negativity towards large group play outside of CW. Welcome to the inevitable backlash from using pubbies for punching bags.
#906
Posted 22 January 2016 - 07:43 PM
Fastwind, on 09 January 2016 - 03:34 PM, said:
bring back the old elo per mech and the old gamemode select systems for far superior enjoyable fun matches
as the solo queue is right now,it's simply a very poor match quality and not much fun
i didn't mind the wait times,for far more enjoyable matches
also CW matching pug's vs grp's didn't work before wont work ever
you already tested it,it failed
why do it again?
We could just simply eliminate mode voting - that would solve skirmish mode practically all the time.
And de-link match score from PSR. Match score is not a measure of skill. Things such as being the first to target enemy mech (scouting bonus) adds to match score - but is no measure of skill. And second, there are players such as myself who have reached their limit of skill - yet the PSR system assumes that more playing the skill will always.
Interestingly, when PSR was first introduced - for about 2 or so weeks - the matches were balanced and good. Thereafter, the MM/PSR pretty much went down the drain.
#907
Posted 23 January 2016 - 08:57 AM
#908
Posted 23 January 2016 - 12:24 PM
I would love to see a true lobby system for group play as the LFG is sort of light at the moment. Like a private match setup perhaps but open to everyone to slot in and go. Have a list of maps open for players to slot in so you can choose a build for that map even. Once sides are set it auto launches. Sort of like stacking for CW attack/defend but you see all the players. 12 man looking for another 12 man to play? No problem there is one waiting on Forest Colony etc. Player skill matching gets hosed there but still.. Might be fun. Quick action button can just spin the MM wheel and jump in just like now. Others have suggested these types of things as well.
This coming from a player who plays solo or groups with 2-3 others MAX all the time. Sure getting rolled by 10-12 man groups is not great but the choices we have for going it alone or with up to 11 friends on the same team is one of the great features of MWO. Does MM need work? Certainly. However going back to hard limits will not help but only hurt matters.
Solo play feels pretty good at the moment to me as most matches feel competitive but this is all general perception and definitely not scientific. Like others I really want CW to have more depth and meaning and agree that when it gets better many large groups will play there.
Cheers.
#909
Posted 25 January 2016 - 01:12 PM
Commander A9, on 07 January 2016 - 09:00 PM, said:
Standard cue matchmaking needs to have voting system removed, unless it's addressing the maps, not the game mode type; game mode type filter needs to be given back to the players. No reason to have it removed. Then you can actually make use of cap accelerators in dedicated modes and not skirmishes.
I said I would not play before because of I didn't want to play conquest, but I got use to conquest now. However you do bring a good point about cap accelerators, what's the point of them in skirmish?
#910
Posted 25 January 2016 - 06:40 PM
All it takes is one or two people to go leroy jenkins and then you are down a hundred tonnes. You 'll notice in matches their are 2-3 players per team who do excessive damage, these are usually the mismatched games. When you have a game where people are posting 4-5 X their tonnage in damage, that you have one of those really good games that are down to the last few mechs, that matchmaker has worked properly.
As for CW ,it comes down to the facts that a 12 man group is probably dropping with a distinct gameplan in mind,specific builds(meta usually), group mindset and good use of voip will tear apart an opposing team that isn't(no surprise there). This is where PGI has to make the CW balanced.
In order to do this CW should ,next to never, have a 12 man vs a motley assortment of solo players. Large groups should be playing against other large groups to give the best chances of a decent matchup. If this isnt possible because of not enough player base then Limiting the number of players in group drops is one way. Maximum of 4-8 players to a group drop with solo players to round out numbers.You could limit this by Xgroup/unit TAGS in a 12 man.This will prevent drops of multiple smaller groups(3 x 4 man etc)on the same map. This will "even out matchmaking "hopefully.
Another CW solution is to allow SOLO players to have prebuilt dropdecks they can pick to drop with. These mechs can be mastered out already allowing people to just "Jump INTO CW" without having to grind out 3 of everything to build a CW dropdeck of their own.This would allow players quicker access AND increase CW player base while dropping players of similar skillsets on opposing teams.It would also be easier for PGI to match up players when 1/3 to 2/3 are in prebuilt mechs with IDENTICAL stats and loadouts. MWO is a very much a game of grinding and spending time to get your mech xpd out and would allow those with limited time to still get into the game.
You could even make an event out of it by having players submit their favorite dropdeck builds and have the community at large vote for the top 3 decks for IS and CLAN. Why try and build dropdecks yourselves when we would do it willingly? We have an excellent community who are devoted to the genre . We' re here, use us to help game development. AND we are willing to do it for some baubles, c-bills( maybe even a hero mech or two ) and some recognition of our contribution. PGI may only have a dev team of less than 50 people ,but MWO, we've got a dev team of thousands. Remember... Play The Game .. Don't Let The Game Play You! O7 MechWarriors!
Edited by Bhodi Li773, 25 January 2016 - 07:03 PM.
#911
Posted 01 February 2016 - 12:10 PM
By this I mean take the pilot tiers and cut them all in half. Tier 1 becomes 1 and 2. Tier 5 becomes 9 and 10. Leave matchmaking at "(Your Tier)+2 OR -2" or however the coding is written. Now I have absolutely no clue how this would change the wait times, but it could offer a more accurate match up. (Someone with the data could project that - I am not going to suggest how, my programming skills are limited to PLCs).
The reason for my suggestion is based entirely on the fact that I recently broke into Tier 3 (only been playing since the Update that introduced Tier 5 [pre-Steam]) and I only solo PUG. The difference is freakin' huge. There is no gradient to slowly work into it. My options now are to play a good match in Tier 4, go to Tier 3 and then lose 25 matches so that I can win one in Tier 4 again. Rinse, repeat. Losing is not any fun, so why play a game if it is not fun? The armchair generals can say "get better or go home, noob" all they want and I know my personality enough to say "%#@& you, peace out."
#912
Posted 07 February 2016 - 04:26 PM
All these new players and players coming from steam are all playing in tier 5-4 where their enemy are all passive and sit still for you where they rack up a lot of high damage numbers and think that's how the game is always going to be. Even in tier 3 most likely will that be the case.
However beyond that they get placed in more experienced players games and get completely destroyed because those players won't sit still for them and they know how to effectively kill mechs. As such their performance greatly drops and start getting wiped game after game because everything they learned from tier 5-4 and most likely tier 3 has no baring on tier 2 and tier 1 games.
If the tier system was based on skill placement not only would that be giving people better placed matches but those new players would be going into different tiers that has a more gradual skill increase where they can better gradually learn how to play the game without being in complete feel good tier ranks for the first few ranks than get completely wrecked when they get placed in the tiers beyond that.
*EDIT* Also reducing group sizes to 4 man's would also be a better experience for quick play, quick play should be the casual gamemode where players can grind mechs and cbills and give a better experience for new players overall. Leaving Community Warfare as actual end game content designed around unit's competing against other unit's and groups for planetary rewards.
Quick Play = Casual gamemode
Community Warfare = non-casual gamemode
Solaris and Tournaments supported by private lobbies = competitive gamemode.
Every type of player gets a gamemode that best supports what they want from the game.
Edited by DarklightCA, 07 February 2016 - 07:47 PM.
#913
Posted 16 February 2016 - 03:21 PM
Its infuriating. Thing is though the t2 losses are pretty much all stomps as if im getting dropped with t5 players against t1. Something is very broken.
#914
Posted 16 February 2016 - 07:00 PM
Otherwise, just keep doing hat you're doing cause you can't please everyone. As long as I can drop when I feel like playing, I'm good.
#915
Posted 23 February 2016 - 12:01 PM
#916
Posted 23 February 2016 - 08:20 PM
Not even getting started on how the Matchmaker has been trying to ram cactus up my rear end,. after I blew through tier 1 and 2 pilot tiers to be stopped in my tracks by these issues and not pilot error or my own costly mistake sucks.damn yous.
#917
Posted 24 February 2016 - 08:28 AM
Group que: honestly, the only problem I see here, is tonnage gets too light after 8+ people, please allow more tonnage for teams with 9 or more pilots.
Since a good portion of the community might complain about this, how about making psr weight heavier the more players they have past groups of 8? What I mean is, groups of 8 or larger will match up with opponents tougher then their current tier (with a maximum tier rating of 1).
#918
Posted 24 February 2016 - 10:32 AM
Lazor Sharp, on 09 September 2015 - 11:45 AM, said:
And with all the 2-5 mans running around, it makes it a night mare for the MM to place all those small groups..
When if you can increase the number of players units field each match, and cut down on the numbers of 2-3 mans, the MM can find matches easier..... and these former 2-3 mans are now not getting stomped so much (Stomps are gonna happen no matter what)
I don't know..... just seams common sense, that a team based game should be played with the most players in units fielding as many players per match as possible, not the other way around.......
There are some nights where I start out as a two man and we do just fine, rolling along with the top dogs in the 8 and 10 mans.
Then We add 1 or 2 more and it all goes to hell. It's also somewhat weird that even when it goes to hell looking at my team of 3 or 4 all of us are above the average damage on the team and even earning kills, but still the 8 or 9 others are terribad and less than 100 damage each.
Considering that in the ideal match you should be doing 250 damage and earning 1 kill each to account for twistings/skilled play plus good CT focus That's sad. whenever you are under 100 damage dealt you are asking your team to make up for a whole person, because a) you are gone and you didn't even do the min threshold of damage to open up any of the opposition.
I was puzzled how this was happening so I've been paying attention to pilot tiers for the folks that join us. The most common early groupers with me are also tier 2 and tier 1 players of some small skill themselves. The later joiners are often recruits or recent joiners, some as low as tier 5. Had a recruit beg off dropping with us because we were seeing SJR and Emp and 228 on the regular and he could not keep up. Always below 100 damage. His tier was 5 and he was totally outclassed.
No idea how the matchmaker weights the various tiers in a group of people, but my assumption was that it would be averaging the tier rounding down. You've got ratings = 8 for that group. 8/3 =2.667 should be Tier 3 average. We saw mostly tier 1 competition that whole time. So either the group queue is starved for players (entirely possible) or the MM averaging needs further adjustment?
#919
Posted 24 February 2016 - 12:38 PM
I like the idea of limiting group sizes to 4-8-12. Math OP!
The matchmaker should also try to give both sides even groupings. Like skill, it will not always be able to satisfy this constraint. They'd have to do internal statistical analysis to determine the impact on w/l of different factors for the solver.
And everyone should be able to do private practice / goofing around matches while waiting. Pretty obvious quality of life improvement really. These don't need to be server authoritative to reduce server load. If that's too complicated, turn off damage to let players do something while waiting.
Edited by 4EVR, 24 February 2016 - 12:43 PM.
#920
Posted 25 February 2016 - 09:33 PM
my situation: i began to play MWO quite recently (1 month more or less), played battletech pnp and all mw games since mech2... for now I currently play either solo or with 2-3 friends that played before me. Got to tier 3 pretty quickly, saw no real change between different tier...
What I see both in solo play or group play, it is like tossing a coin:
-first possible outcome: Win 12:2 or 12:4...good to great score each and stomp the other team. Other players on our side where usefull/worked together/direwolf on our side did not hide behind a rock when a locust used machine gun on it and instead exploded it with an alpha strike...
-second possible outcome: Loose 2:12 or 4:12 ... lame score... died stupidly because everyone hid behind a rock or ran in circle trying to run away when a locust sprayed machine gun on them leaving big fatlas and king crab alone for the picking by other team... the rest of the story should be predictable...
On the CW side... it is either like the above coin tossing or an execution by a 12 man team.
I didn't see how it was before so i'm not well placed to propose anything, but I know there is a problem somewhere ;P
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users